Fingerprint recognition is one of the most widely used biometrics in the access control industry.This is because fingerprints are one of the oldest forms of personal identification, inexpensive to collect and analyze, and they are stable. Biometric fingerprint access control applications use one or both of two fingerprint characteristics: ridge patterns and minutiae details, which are unique features found within the patterns. Some extremely hi-tech biometric fingerprint scanners not only require a fingerprint to match, but they employ temperature and humidity monitors to ensure that a live finger is being scanned.
There are three basic patterns, of fingerprint recognition ridges: the arch, loop, and whorl.
- The arch is made up of ridges lying one above the other in a general arching formation.
- The loop is made up of ridges that enter from one side of the finger, form a curve, and then exit on the same side.
- The whorl is made up of ridges that form a circular pattern around a central point.
Minutiae recognition, as seen in Figure 3-19, is the most common form of biometric access control fingerprint identification. Minutiae include the discontinuities that interrupt the otherwise smooth flow of fingerprint ridges and the abrupt ridge endings and bifurcations. The major minutia features of fingerprints are bifurcation, dots, and ridge endings.
• Bifurcations are the points where one ridge divides into another.
• Dots are very small ridges, no longer than the width of adjacent ridges.
• Ridge ending is the point where a ridge terminates.
The ridge patterns and minutiae are important in fingerprint analysis since no two fingerprints are identical.
A fingerprint recognition sensor is a device used to capture a digital image of a fingerprint. There are several commercially available sensor technologies for electronically collecting fingerprints, including optical, silicon, ultrasound, and light emitting.
Optical technology is the oldest and most widely used collection method. For optical collection, the finger is placed on a proprietary, coated platen. A charged coupled device (CCD) converts the image of the fingerprint, dark ridges and light valleys, into a digital format. Optical devices can withstand temperature fluctuations, are inexpensive, and provide resolution typically ranging from 500 to 1,000 dots per inch (dpi). Drawbacks of optical technology primarily relate to the requirement to make contact with the reading device’s surface. Users leave residual fingerprints, grease, dirt, and other materials that interfere with the reader’s performance and cause the reading surface to wear. Some of the new contactless optical fingerprinting detectors eliminate these deficiencies.
Silicon chip technology, or capacitance technology, has gained considerable acceptance since its introduction to the marketplace in the 1990s. In most chip systems, the sensor acts as one plate of a capacitor and the finger acts as the other. The capacitance between the detector and the finger is converted into an 8-bit grayscale digital image. Silicon generally produces better image quality with less surface area than optical technology. Silicon-based equipment can be much smaller than optical devices. The primary drawbacks are a shorter track record for durability in suboptimal environments compared to optical technology, the smaller scanning surfaces, and issues relating to the requirement to touch the reader’s surface.
Ultrasound technology has only recently been introduced to the biometric market, thus it is not widely used.
Ultrasound uses high frequency sound waves to measure the impedance of the finger, air, and platen to generate a signal. Sound waves penetrate the dirt, grease, and other contaminates to obtain an image of the tissue and veins of the finger. This process can obtain usable prints in some situations that would impede optical systems. Some systems can resolve fingerprints of small children, petite adults, and persons with dry, rough, or worn fingers that can be difficult for other systems. Systems can also differentiate between dead skin and live skin. The technology’s main drawback is its newness in the market and lack of a track record.
Light emitting sensor technology uses electroluminescence film on its sensors. When a finger is in contact with the film and an electric field is applied, a high resolution image of a fingerprint is produced. The film is durable, works in direct sunlight, and is very lightweight, making it easily adaptable for mobile applications.
Regardless of the technology applied, there are several steps required to convert a high-quality captured print into a compact template. Feature extraction, as this process is called, is the basis by which fingerprint technology produces a usable sample. The exact processes used are proprietary to each access control vendor. the user places a finger or fingers on the reader. The reader captures the image and then, based on its design, will use complex algorithms to characterize the print for comparison to the template in the access control database.
Fingerprint scanning technology continues to advance in multiple ways, including image capture unaffected by lighting environment, spoofing detection of fake fingerprints, and liveness detection. Liveness detection ensures a pulse is associated with the captured fingerprint sample to prevent the exploitative use of dismembered digits or fake digits. Also, highly detailed 3-D representations of finger surfaces and contactless live scan devices are being developed that produce prints of fingers, and in some cases, the entire hand, from a standoff position.
Fingerprint-based access control systems are used for both identification and verification purposes. The technology is mature and available from a variety of vendor sources. Silicon chip and light emitting sensor based systems can be small enough to use with mobile devices, such as smartphones, tablets, or laptop computers.
Fingerprint-based access control systems are for indoor/outdoor use and respond well to a wide range of environmental humidity and temperature conditions.
These systems are not recommended for applications in which users wear gloves or work in conditions that cause abrasion to the fingers and hands. Such applications might include construction sites, nuclear/chemical surface contamination areas, or foundries.
The most fundamental feature of a reader is the size of the scanner’s sensing area. The average size of a fingerprint is approximately 0.5 by 0.7 inches (smaller for children and females, larger for males). A smaller reader surface obtains a partial print, which may not include all of the area required for enrollment scans. Smaller reader surfaces render the system more sensitive to differences in finger positioning for each reading and may result in unacceptable false rejection rates. Optical and ultrasound systems are more likely to have larger sensing areas than chip systems, because large chips that perform uniformly over the entire surface are difficult and expensive to make. The sensing area of a chip system is dependent on the size of the chip. Rates of failure to enroll can be high for some systems when the user population includes children, petite adults, or persons with finger surfaces that are worn, rough, or dry. Recent technologies claim to have improved performance for these difficult categories.
Another important parameter is the number of pixels characterizing an image (i.e., resolution). A minimum image resolution of 500 dpi is required by FBI-compliant systems. To capture minutiae requires a minimum of 250 to 300 dpi. Resolutions higher than 500 dpi may be necessary to enroll children and petite adults whose print characteristics are too small and close for some systems to read.
Finger placement is very important in order to obtain an accurate and repeatable print. Other parameters of importance include the cleanliness of the finger, the contrast, and the amount of geometric distortion present in the sample. Systems that require contact with the reading surface require relatively frequent cleaning to remove latent prints and maintain high sensing accuracy.
Some fingerprint systems may be vulnerable to spoofing using 3-D molds of an enrollee’s fingerprint. Technology to verify that a finger is live is available in some systems to reduce the likelihood of spoofing.
- Access Control Technologies Handbook
see also :